Tracheal occlusion modulates the gene expression profile of the medial thalamus in anesthetized rats

Abstract

Conscious awareness of breathing requires the activation of higher brain centers and is believed to be a neural gated process. The thalamus could be responsible for the gating of respiratory sensory information to the cortex. It was reasoned that if the thalamus is the neural gate, then tracheal obstructions will modulate the gene expression profile of the thalamus. Anesthetized rats were instrumented with an inflatable cuff sutured around the trachea. The cuff was inflated to obstruct 2–4 breaths, then deflated for a minimum of 15 breaths. Obstructions were repeated for 10 min followed by immediate dissection of the medial thalamus. Following the occlusion protocol, 588 genes were found to be altered (P < 0.05; log₂ fold change ≥ 0.4), with 327 genes downregulated and 261 genes upregulated. A significant upregulation of the serotonin HTR2A receptor and significant downregulation of the dopamine DRD1 receptor genes were found. A pathway analysis was performed that targeted serotonin and dopamine receptor pathways. The mitogen-activated protein kinase 1 (MAPK1) gene was significantly downregulated. MAPK1 is an inhibitory regulator of HTR2A and facilitatory regulator for DRD1. Downregulation of MAPK1 may be related to the significant upregulation of HTR2A and downregulation of DRD1, suggesting an interaction in the medial thalamus serotonin-dopamine pathway elicited by airway obstruction. These results demonstrate an immediate change in gene expression in thalamic arousal, fear, anxiety motivation-related serotonin and dopamine receptors in response to airway obstruction. The results support the hypothesis that the thalamus is a component in the respiratory mechanosensory neural pathway.

the respiratory system is continually active, and any prolonged interruption is a threat to an animal's survival. Thus it is of critical importance to maintain ventilation in the face of a variety of stimuli by adjusting the breathing pattern. Clark and von Euler (13) described the relationship between lung volume and breath timing in anesthetized cats. They demonstrated that inspiratory time (Ti) depends on inspiratory volume and that the subsequent expiratory time (Te) depends on the preceding Ti. A similar volume-timing relationship was found when a mechanical stimulus in the form of an external resistive load was applied. Loading of the inspiratory (57) or expiratory (33) phase caused a decrease in volume and an increase in the duration of the respective loaded breath phase. The response to the added loads was called the respiratory load compensation reflex. Load compensation is a sensory motor response.

A sufficiently high load on the respiratory system results in the sensation of dyspnea, or breathlessness, which is one of the primary symptoms in pulmonary and cardiovascular diseases (40, 44). It is one of the main symptoms that limits patients with obstructive pulmonary diseases, and it can also be considered one of the most important factors in determining the severity of the disease and the health-related quality of life (11). Sensory information of breathing is continuously sent to the respiratory centers in the brain stem; minor changes in breathing are controlled by these centers without the activation of higher brain centers. When changes in respiratory information reach a certain threshold, then a gating-in process occurs, resulting in the cognitive awareness of breathing (2, 44). The conscious awareness of breathing as in dyspnea requires the activation of higher brain centers (44).

The proposed brain structure involved in the gating of respiratory sensory information is the medial thalamus. The thalamus is the largest structure in the diencephalon, located centrally in the brain. It is ideally situated to receive incoming sensory information and to send nerve fibers out to higher brain centers. Midline thalamic nuclei receive projections from areas such as the periaqueductal gray (36), the parabrachial nucleus (35), the superior colliculus (37), and the brain stem (34). The thalamus integrates many bidirectional connections with several regions of the cortex, most importantly the recurrent loop to and from various cortical areas, but also to the amygdala (for emotional processing), the hippocampus (for learning and memory), and the limbic system (32, 42a). A single thalamic nucleus can send efferents to multiple cortical areas (26). Intralaminar thalamic nuclei have diffuse projections to the cerebral cortex (31). Reciprocal corticothalamic neurons connect to the thalamic relay and interneurons (41). It is believed that this corticothalamic feedback functions in “egocentric selection,” which refers to the ability of cortical neurons to analyze thalamic input, select certain sensory features, and then amplify the transmission of these features by feedback to the thalamus (49). This suggests that cortical areas can either enhance or suppress specific information. Other top-down connections such as from the cingulated gyrus and prefrontal cortex feeding back onto the thalamic neurons help in selecting stimuli that are relevant, salient, and novel (32). In some instances, the thalamus may impair rather than facilitate the processing of environmental stimuli, such as during trauma (32). Therefore, functionally, the thalamus is thought to serve as the processing and relay station that most sensory information must pass before reaching the cortex (32, 41, 47). For the well studied visual, auditory, and somatosensory systems, different thalamic relay neurons are responsible for relaying modality-specific information. Visual stimuli pass through the lateral geniculate nucleus, auditory information pass through the medial geniculate body, and somatosensory stimuli are processed by the ventrobasal complex (1). Previous studies using intrinsic, transient tracheal occlusions (ITTO) in our laboratory have demonstrated increased c-Fos expression in response to tracheal occlusions in the medial thalamus, indicating activity of neurons in this brain area due to the respiratory stimuli (46, 53). Thus it was hypothesized that respiratory sensory information may be processed and relayed to higher brain centers through the medial thalamus.

Thalamic relay neurons receive both excitatory (glutamatergic) and inhibitory (GABAergic) inputs, and the balance between these transmission signals is what eventually determines the response (41). Modulations of any component in the neural transmission pathway (such as transmitters, receptors, or transporters) alter the signal processing of the thalamic relay neurons and thus change sensory information gating. Specifically, modulations to the serotonergic system appear to be important in gating processes. For example, activation of the serotonin receptor subtype 2A (HTR2A) reduces sensory gating so that more sensory information reaches cortical areas and thus consciousness (29).

In the present study, the load compensation response to ITTO was determined and comparative microarray analysis was performed to examine the molecular changes that occur immediately after ITTO. It was hypothesized that ITTO would result in a load compensation response consisting of an increase in breath phase time and a more negative esophageal pressure during occlusions only. In addition, it was hypothesized that ITTO would induce short-latency (<10 min) gene expression changes in the medial thalamus, specifically, changes in components of neurotransmitter systems.

MATERIALS AND METHODS

Animals.

Eight male Sprague-Dawley rats (276.8 ± 47.5 g) were housed two per cage in a temperature-controlled room (72°F) on a 12:12-h light-dark cycle and with free access to food and water. All animal experiments were approved by the Institutional Animal Care and Use Committee of the University of Florida.

Surgical procedures.

Animals were anesthetized by intraperitoneal injection of urethane (1.3–1.5 g/kg), and anesthesia was supplemented as needed (20 mg/ml). Anesthetic depth was verified by the absence of a withdrawal reflex to a rear paw pinch. Body temperature was measured using a rectal probe and maintained at 38°C with a heating pad. Animals were spontaneously breathing room air.

Esophageal pressure (Pes) was measured by inserting one end of a saline-filled tube through the mouth into the esophagus. The other end of the tube was connected to a polygraph system (model 7400, Grass Instruments) via a pressure transducer, and the analog output was amplified, digitized at 1 kHz (CED model 1401, Cambridge Electronics Design), and computer processed (Spike2, Cambridge Electronics Design). Pleural pressure changes were inferred from relative changes in Pes.

Diaphragm electromyographic (EMG_dia) electrodes were prepared from Teflon-coated wire. The ends of the wires were bared, bent, and hooked into the costal diaphragm through a small incision in the abdominal skin. Two electrodes were inserted for bipolar EMG_dia recordings. The electrode wires were connected to a high-impedance probe. The signal was amplified (P511, Grass instruments) and band-pass filtered (30–300 Hz). The analog outputs were digitized and processed as described above.

The trachea was exposed through a ventral incision via blunt dissection of surrounding tissue. An expandable cuff was sutured around the trachea, two cartilage rings caudal to the larynx. The cuff was connected to an air-filled syringe via a thin rubber tube. The syringe was used to inflate and deflate the cuff bladder. Before the experiments, the inflation pressure needed to completely compress the trachea was determined using an excised trachea. A complete compression occluded the airway during both inspiration and expiration. Deflation restored the trachea back to its original condition to allow unobstructed breathing.

Experimental protocol.

The experimental group (n = 4) was allowed to breathe unobstructed for 60 min following surgical preparations. Then the cuff was inflated to occlude the trachea for 2–4 breaths, followed by deflation of the cuff for a minimum of 15 breaths. The occlusions were repeated for a total of 10 min. Pes and EMG_dia were monitored continuously throughout the experiment to verify onset and removal of tracheal occlusions. The control animals (n = 4) underwent the same surgical procedure, 70 min of unobstructed breathing, but did not receive tracheal occlusions. Immediately after completion of the 70-min postsurgical period, the animals were decapitated, and their brains were removed. The medial thalamus was located, excised, frozen in liquid nitrogen, and stored at −80°C until further use.

Physiological data analysis.

Data were analyzed offline using analysis software (Spike 2, Cambridge Electronics Design). The EMG_dia was rectified and integrated (time constant = 50 ms), and inspiratory time (Ti), expiratory time (Te), and total time (Ttot) for each breath were calculated from the integrated EMG_dia signals. Ti was measured from the onset of the inspiration-associated increase in EMG_dia activity to the point at which EMG_dia peak activity began to decline (Fig. 1). Te was measured from the end of Ti to the onset of the following inspiration (Fig. 1). Baseline EMG_dia was defined as the minimum value of the EMG_dia during expiration. The EMG_dia amplitude (ΔEMG_dia) was calculated as the difference between baseline and peak EMG_dia. Pes amplitude was calculated as the difference between baseline and peak Pes. For the experimental group, within each occlusion presentation, the control breath (C) was defined as the complete breath immediately before occlusion (O) application, and the recovery breath (R) was defined as the first complete breath immediately after termination of occlusion. For the 10-min occlusion trial, one control breath, two occluded breaths (O1 and O2), and one recovery breath were measured for each occlusion presentation. Data from at least 28 occlusion trials for each experimental animal were used for analysis. For the control group, breathing pattern of control breaths was measured at matched time points.

Fig. 1.

Ventilatory response to intrinsic, transient tracheal occlusions (ITTO). The top bar represents time period of an occlusion. Control (C), first occlusion (O1), second occlusion (O2), and recovery (R) breaths are indicated. The determination of inspiratory time (Ti) and expiratory time (Te) are shown. Breath timing (Ttot = Ti + Te) increased mostly due to an increase in Te. The esophageal pressure (Pes; top trace) was more negative during O1 and O2, then returned to control immediately after the occlusion was removed.

SigmaStat for Windows version 3.5 (Systat Software) was used for all statistical analyses of physiological data. All values are reported as means ± SD. Ti, Te, Ttot, Pes, and ΔEMG_dia for C breaths were compared between experimental and control animals using one-way ANOVA. Comparisons between C, O, and R breaths in the experimental animals were analyzed using one-way repeated-measures ANOVA.

Total RNA isolation.

Total RNA was isolated from medial thalamic tissue with RNA Stat-60 (Tel-test, Friendswood, TX). The frozen tissue (10–20 mg) was homogenized in Stat-60, and chloroform was added. The mixture was vortexed for 15 s and centrifuged at 12,000 g for 15 min at 4°C. The upper aqueous phase containing the RNA was carefully extracted. The extraction step was repeated, and the RNA was precipitated with isopropanol. Following another centrifugation at 12,000 g for 40 min at 4°C, the pellet was washed twice with 80% ethanol and air dried. To inactivate RNases, the pellet was resuspended in 40 μl RNA secure (Ambion, Austin, TX), following the manufacturer's protocol. A total of 10 μg of RNA was treated with DNase to avoid contaminating DNA using DNA-free (Ambion), following the manufacturer's protocol. The quality of total RNA was assessed with the Agilent 2100 BioAnalyzer (Agilent Technologies, Palo Alto, CA), and the quantity was determined on a NanoDrop spectrophotometer (NanoDrop Technologies, Wilmington, DE).

RNA amplification and microarray analysis.

Array hybridizations were performed using a two-color reference design; each of the eight thalamic samples (4 experimental and 4 control animals) was hybridized against a common reference sample (Fig. 2). The reference material was obtained by pooling all eight samples. The cDNA synthesis, cRNA labeling, and hybridization were performed following the manufacturer's kits and protocols (Agilent Low RNA Input Fluorescent Linear Amplification Kit and Agilent 60-mer oligo microarray processing protocol; Agilent, Palo Alto, CA). The thalamic experimental and control samples were labeled with Cy5, whereas the reference sample was labeled with Cy3. The cRNA was amplified and purified using the QIAGEN RNeasy Kit (Qiagen). Dye incorporation was determined by using Nanodrop (>13 pmol/μg RNA). Each of the arrays (4 × 44,000 rat genome DNA oligo microarrays, Agilent Technologies, Amadid: 014879) was then loaded with the reference mixture (825 ng) and either experimental or control RNA (825 ng) (Fig. 2). Hybridization was carried out in a microarray hybridization chamber at 65°C for 17 h. The glass slides were then washed and scanned with a laser-based detection system (Agilent). Text files of the raw data from this study have been deposited in NCBI's Gene Expression Omnibus (18) and are accessible through GEO Series accession number GSE25153 (http://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE25153).

Fig. 2.

Schematic representation of microarray analysis using a two-color reference design. Each of the eight arrays was loaded with the reference mixture and either experimental or control RNA. There were eight arrays with 44K probe; eight animals with experimental animals (n = 4) and control animals (n = 4), with1 pooled reference sample; the experimental and control samples were randomized on the microarrays.

A log₂ transformed signal ratio between the experimental channel and the reference channel was calculated for each spot, followed by within-array lowess transformation and between array scale normalization on median intensities (56). t-Test for the samples was performed on normalized log₂ transformed signal ratios of each probe individually, followed by multiple testing correction using a Benjamini-Hochberg approach (5). Genes were considered differentially expressed if the P value was ≤0.05 and the log₂ fold change was ≥0.4.

Gene ontology and pathway analysis.

Gene ontology annotations were derived from similarity searches of the NCBI Gene database. A blastn search for each of the 44,000 probes was performed to retrieve the gene ontology (GO) annotation. Once the GO annotations were retrieved, a GO tree was built following the hierarchical structure for the whole array. Then, another GO tree for the significant regulated genes was built. The two trees were compared at each node by running a Fisher's exact test (P ≤ 0.05) when traversing the tree branches. Significantly overrepresented GO categories were identified by the Fisher's P value, and the false discovery rate was determined.

Some of the genes that showed significant modulation were scanned against the Pathway Studio ResNet database (Ariadne Genomics, Rockville, MD). This database uses published information and catalogs the relationships between biological entities. Pathway Studio (Ariadne Genomics) was used to identify and graphically display the functional interactions between the selected genes (43).

RESULTS

Physiological responses to ITTO.

Breath timing and Pes response to ITTO are shown in Fig. 1. Breathing frequency slowed due to an increase in Te. The Pes was more negative during occlusion and returned to baseline immediately after termination of occlusion. Comparisons of control breaths between experimental and control animals demonstrated no significant differences in Ti, Te, Ttot, Pes, or EMG_dia (Table 1).

Table 1.

Control breath comparisons

	Experimental Group	Control Group	P Value
Ti, s	0.228 ± 0.014	0.228 ± 0.014	0.997
Te, s	0.488 ± 0.104	0.411 ± 0.073	0.277
Ttot, s	0.715 ± 0.115	0.639 ± 0.073	0.306
Pes, V	−0.038 ± 0.013	−0.038 ± 0.011	0.967
ΔEMGdia, AU	0.697 ± 0.583	0.444 ± 0.246	0.454

Control breath comparisons of breath phase timing, esophageal pressure (Pes), and diaphragm electromyographic amplitude (ΔEMG_dia) between experimental and control animals. Control breaths were defined as the last complete breath immediately before an occlusion for experimental animals and at matched time points for controls animals. Ti, inspiratory time; Te, expiratory time; Ttot, total time. Values are means ± SD. P values are from one-way ANOVA.

Comparisons between control, occluded, and recovery breaths within each experimental animal revealed a load-compensating reflex (Table 2; Fig. 3). Te, Ttot, and Pes were significantly different between C and O, and between O and R, but not between C and R. There was a trend toward significance in Ti (Table 2). EMG_dia showed no significant differences.

Table 2.

Comparisons between control, occlusion, and recovery breaths in the experimental animals

	Control	O1 + O2	Recovery	P Value
Ti	0.228 ± 0.014	0.248 ± 0.014	0.211 ± 0.023	0.063
Te	0.488 ± 0.104	0.615 ± 0.172*,†	0.472 ± 0.128	0.011
Ttot	0.715 ± 0.115	0.864 ± 0.178*,†	0.683 ± 0.145	0.010
Pes	−0.038 ± 0.013	−0.055 ± 0.012*,†	−0.041 ± 0.015	0.002
ΔEMGdia	0.697 ± 0.583	0.736 ± 0.591	0.717 ± 0.605	0.197

Comparisons between control, occlusion (O1 + O2), and recovery breaths in the experimental animals with combined values from O1 and O2. Values are means ± SD.

^*Significantly different from recovery breath (P < 0.05; one-way, repeated-measures ANOVA).

^†Significantly different from control breath (P < 0.05; one-way, repeated-measures ANOVA).

Fig. 3.

Group mean ventilatory response to ITTO for experimental animals. Comparisons between control (C), occlusion (O), and recovery (R) breaths. A: inspiratory duration (Ti). B: expiratory duration (Te). C: total breath duration (Ttot). D: esophageal pressure (Pes). E: diaphragm EMG (ΔEMG_dia) amplitude normalized by dividing the O1, O2, and R breaths by the C breath relative ΔEMG_dia. Values are means + SD.

Modulation of gene expression profile following ITTO.

Changes in gene expression profiles in the medial thalami of rats were assessed by cDNA microarray analysis. Statistical analysis of the microarray data showed that a total of 588 genes were altered (P < 0.05; log₂ fold change of ≥0.4) following the occlusion protocol compared with the control group, with 327 genes downregulated and 261 genes upregulated (see supplementary data available online at the Journal of Applied Physiology website). Some candidate genes of interest included genes involved in stress-related pathways (Table 3).

Table 3.

Genes of interest that were significantly regulated following ITTO

Log Fold Change	P Value	Gene Symbol	Description
+1.30	0.0057	PLAU	Plasminogen activator, urokinase
+1.07	0.0425	HTR2A	Serotonin receptor 2A
+0.91	0.0060	TNFRSF14	Tumor necrosis factor receptor superfamily
+0.75	0.0098	CHRNB1	Cholinergic receptor, nicotinic, beta polypeptide 1
−0.74	0.0316	KCNJ3	Potassium inwardly-rectifying channel
−0.71	0.0045	COX6B2	Cytochrome c oxidase subunit
−0.52	0.0104	DLG4	Discs, large homolog 4
−0.45	0.0128	DRD1A	Dopamine receptor D1A
−0.42	0.0271	PRKAA2	Protein kinase alpha 2 catalytic subunit

ITTO, intrinsic, transient tracheal occlusions.

We further characterized the types of genes altered in the medial thalamus in response to tracheal occlusions using gene ontology categories. Table 4 shows the GO categories for biological processes that were overrepresented among the regulated genes (P < 0.05; FDR < 0.1). The most significantly regulated GO categories were “anti-apoptosis,” “response to stress,” “regulation of enzyme activity,” and “MAPKKK cascade.” Interactions between some of the candidate genes and their effects on neurotransmission were visualized using Pathway Studio software (Fig. 4).

Table 4.

Highly regulated biological processes following ITTO were found with GO analysis

GO ID	GO Name	No. of Genes Selected	No. of Genes on Array	Fisher P Value	FDR
0006916	Anti-apoptosis	25	293	8.6 e−5	0.0497
0006950	Response to stress	93	1780	9.6 e−5	0.0277
0050790	Regulation of enzyme activity	37	575	4.8 e−4	0.0937
0000165	MAPKKK cascade	22	281	6.6 e−4	0.0964
0043066	Negative regulation of apoptosis	28	411	1.0 e−3	0.0998
0043069	Negative regulation of programmed cell death	28	413	1.1 e−3	0.0916

GO, gene ontology.

Fig. 4.

Pathway analysis for dopamine-serotonin receptor gene changes in the medial thalamus with ITTO. Interaction between dopamine (DRD1) and serotonin receptors (HTR2A) under the control of MAPK1. Both DRD1 and HTR2A have actions on the same neurotransmitters and small molecules.

DISCUSSION

ITTO load compensation.

When an animal is challenged with an increase in respiratory mechanical load, the respiratory control system generates the load compensation reflex. This load compensation response has been reported in anesthetized animals using external resistive loads to breathing and is characterized by the recruitment of respiratory muscle activity, an increase in breath duration, and a decrease in tidal volume (6, 8, 13–16, 57). Depending on the timing within the breath phase of the added resistive load (end-inspiratory vs. end-expiratory), inspiratory or expiratory duration is increased. The load compensation reflex is also dependent on selective phase loading (inspiration only or expiration only) and loading that occurs throughout the entire breath. In this study, we applied complete tracheal occlusions for multiple breaths so the load was applied on both the inspiratory and expiratory breath phases. Load compensation breathing pattern was characterized by a more negative Pes and an increase in Ttot, primarily due to an increase in Te, whereasTi increased nonsignificantly. The diaphragm activity (EMG_dia) was not significantly modulated, so the more negative Pes is most likely due to the ITTO respiratory mechanical changes. Inspiratory duration did not change significantly, which is in congruence with an unaltered EMG_dia signal. The load compensation response lasted for as long as the trachea was occluded, and breathing patterns returned to baseline levels immediately after withdrawal of occlusion, demonstrating that tracheal occlusions using an inflatable cuff are reversible.

Airway occlusions induce serotonin receptor HTR2A and reduce dopamine receptor DRD1.

Modulations to the serotonergic and dopaminergic systems appear to be important in gating processes. Activation of the serotonin receptor subtype 2A (HTR2A) reduces sensory gating so that more sensory information reaches cortical areas and thus consciousness, which in turn can lead to the pathology of anxiety disorders (29). In addition, Carlsson and colleagues (10, 17, 24) have postulated that hyperactivity of dopamine reduces the protective influence of the inhibitory action of striatothalamic GABAergic neurons onto thalamocortical glutamatergic neurons, which can then lead to sensory overload and hyperarousal, confusion, or psychosis. Indeed, it has been shown that neuroleptic drugs that block receptors in the brain's dopaminergic pathways improve sensory functioning and gating in schizophrenic patients (20).

Medial thalamic mRNA transcripts of the serotonin receptor HTR2A were upregulated following tracheal occlusions. It is well known that the serotonin system plays an important role in a variety of human psychopathological conditions, particularly mood and anxiety disorders (12, 25). Antidepressant treatment has focused on modulating serotonergic neurotransmission (30). One of the challenges of the serotonin system is the sheer complexity of the 14 known receptor varieties categorized into seven receptor subtypes (28). Specifically, the HTR2A receptor has been identified to be involved in anxiety disorders in dogs (52). Antagonists to this receptor may be useful therapeutic agents in the treatment of generalized anxiety disorder and psychosis (29, 30). Activation of HTR2A receptors weakens the sensory gating so that more sensory information is able to reach consciousness, which in turn can lead to the pathology of anxiety disorders (29). Behavioral studies in HTR2A knockout mice have shown changes in anxiety-related but not depression-related paradigms (54). The knockout mice exhibited greater exploratory and risk behavior in conflict paradigms, such as the open field, dark-light choice, elevated plus-maze, and novelty-suppressed feeding tests. In depression-related behaviors, as measured by the forced swim test and the tail suspension test, the mice did not differ significantly from wild-type mice (54).

The dopaminergic system has been identified to be involved in a myriad of functions, such as motivation, reward, pain processing, learning, and memory (3, 48). Hyperfunction of this system has been hypothesized to be associated with schizophrenia and attention deficit hyperactivity disorder (10, 17, 24). This hypothesis is based on the antagonistic interaction between dopamine and glutamate projecting on GABAergic striatal neurons that exert an inhibitory effect on thalamocortical glutamatergic neurons, thereby filtering out part of the sensory input to the thalamus to protect the cortex from a sensory overload (10). Hyperactivity of dopamine or hypofunction of the corticostriatal glutamate pathway should reduce this protective influence and could thus lead to confusion or psychosis (10, 21). Neuroleptic drugs that block dopaminergic pathways improve sensory functioning and gating in schizophrenic patients (20). Interaction between serotonin and dopamine systems can have either potentiating or antagonizing effects (39). HTR2A antagonists have been shown to increase dopamine release in a variety of brain regions (27). In the present study, there was an immediate upregulation of serotonin HTR2A receptor gene in the medial thalamus and a coincident downregulation of the dopamine DRD1 receptor gene. This suggests that acute ITTO modulates the thalamic serotonin-dopamine systems, consistent with increased fear conditioning and decreased gating, and may be related to respiratory load-related anxiety and/or depression.

Airway obstruction in disease and association with anxiety and depression.

Asthma is a respiratory disease characterized by reversible airways obstruction, airway inflammation, and hyperreactive airways (51). Our animal model of reversible tracheal obstructions, ITTO, thus mimics one component of asthma. It is well acknowledged that respiratory diseases, such as asthma and chronic obstructive pulmonary disease (COPD), are associated with significantly higher rates of anxiety and depression compared with the general population (42). In addition, there are significantly more individuals suffering from a respiratory or lung disease with panic disorder or major depression than individuals without such a respiratory diagnosis (22). In a sample of 189 patients from a Brazilian outpatient clinic for the treatment of asthma and COPD, Carvalho et al. (11) found that a significant number of patients with controlled and uncontrolled asthma exhibited moderate to severe anxiety as determined by the State-Trait Anxiety Inventory (CA 97.5%, UA 93%), and 74% of COPD patients. Depression scores, as measured by the Beck Depression Inventory, were less pronounced in these patients (CA 20%, UA 49%, COPD 29%) but nevertheless a cause for concern. In a similar study on 132 pulmonary disease patients in a Greek hospital, Moussas et al. (42) found that a total of 49.2% showed moderate or severe depression and 26.5% had anxiety. Fernandes et al. (19) demonstrated a positive association of higher degrees of asthma severity with increased anxiety. In this study, 70% of the patients had a clinical diagnosis of anxiety, and anxiety was associated with worse subjective asthma outcomes and increased use of medication and healthcare services. Of 62 asthmatic patients from an outpatient clinic in Brazil, 24.1% had major depression disorder and 33.8% had an anxiety disorder as diagnosed by the Mini-International Neuropsychiatric Interview (51). However, there was no association between the severity of asthma and the prevalence of anxiety and depression (51). In a study on Veteran's Affairs patients with chronic breathing disorders, 50.1% showed moderate to severe depression, and 64.2% had moderate to severe anxiety symptoms (38). In a 20-year longitudinal and cross-sectional study with 591 participants between the ages of 19 and 40 yr, Hasler et al. (23) found that asthma was strongly associated with panic disorder and that the presence of asthma predicted subsequent panic disorder. Patients with severe asthma and a comorbid psychiatric disorder had an almost 11-fold increased risk for two or more asthma exacerbations and an almost 5-fold increased risk for two or more hospitalizations during the past year compared with patients with severe asthma without psychiatric disorder (50). These results show that anxiety and depression are associated with respiratory diseases and that in some vulnerable individuals an increase in anxiety may lead to panic disorders, possible through dyspnea-induced fear conditioning (23). Thus our results suggest that acute ITTO conditioning modulates serotonin-dopamine gene expression in the medial thalamus consistent with reduced gating of respiratory obstruction stimuli, which may be one mechanism for increased anxiety and/or depression in obstructive respiratory diseases.

Airway occlusions alter genes involved in anti-apoptosis.

Gene ontology analysis identified anti-apoptosis, negative regulation of apoptosis, and negative regulation of programmed cell death as being affected by airway occlusions. These findings suggest that the medial thalamus may be increasing cell-protective mechanisms. A postmortem study examining anatomical abnormalities in the thalamus of patients diagnosed with major depressive disorder discovered that in these subjects the mediodorsal nucleus of the thalamus had a significantly increased total number of neurons compared with nonpsychiatric subjects (55). Other studies have demonstrated volume reductions in prefrontal cortex (7) and hippocampus (9) as well as decreased number of glia in the cortex (45). Young et al. (55) suggested that the elevated number of neurons in the medial thalamus may have had this reducing effect due to its projections to these other brain areas because excessive glutamatergic thalamic neurons could lead to excitotoxicity. Alternatively, increased GABAergic neurons could result in decreased output to the cortex, thus reducing the need for glial support. It is not known which neuron population (excitatory projection neuron or inhibitory interneuron) is elevated in major depressive disorder. The results of the present study showing modulated pathways involved in anti-apoptosis could suggest a first step in the development of depression after airway occlusions.

Functional analysis.

Pathway Studio was used to visualize changes of gene expression following ITTO. An interaction between the dopamine receptor DRD1 and the serotonin receptor HTR2A exists that is controlled by MAPK1 (Fig. 4). MAPK1 positively regulates the NMDA receptor, which in turn inhibits DLG4 and acts on DRD1. At the same time, MAPK1 inhibits CAV1, which regulates HTR2A. Following tracheal occlusions, the downregulated expression of MAPK1 could have led to a decreased stimulation of the NMDA receptor and a decreased inhibition of CAV1. Even though NMDA receptor and CAV1 did not show differential gene expression, MAPK1 could potentially regulate the function of these genes. The decreased function of NMDA receptor would lead to a decrease in DRD1; indeed, a significant decrease in DRD1 expression was found in the occlusion group. Decreased NMDA receptor function could also result in less inhibition of DLG4, which in turn could result in increased stimulation of HTR2A. Furthermore, decreased inhibition of CAV1 means increased function and increased stimulation of HTR2A. HTR2A gene expression was indeed upregulated following occlusions. DRD1 and HTR2A control the activation and release of neurotransmitters and other small molecules. Thus ITTO elicited a reciprocal interaction between thalamic DRD1 and HTR2A that seems to be predominantly regulated by MAPK1.

Methodological considerations and future studies

Specificity of medial thalamic nuclei.

The medial thalamus consists of multiple nuclei, midline and intralaminar nuclei, that may have different specific functions. Furthermore, the medial thalamus contains a variety of anatomically and functionally heterogeneous neurons (4), such as excitatory and inhibitory interneurons. The thalamic microarray sample in this study was not targeted to a specific nucleus or group of neurons and rather to the thalamus in general, thus gene expression may be multifactorial. Future studies need to be carried out to determine the response of specific thalamic neurons and nuclei comprising the midline and intralaminar nuclei and the connectivity between other brain areas and these neurons.

Genomics vs. proteomics.

The use of microarray technology is a method to study changes in gene expression in cells and following specific stimulus interventions. However, a change in mRNA levels of a particular gene does not necessarily translate into the same change in protein levels of that gene. Most microarray studies that use protein validation show modulation of mRNA and protein in the same direction and with similar fold changes. Some studies, however, have demonstrated distinctly different fold changes or even opposite regulation of mRNA and the associated protein. The advantage of gene microarray studies is the identification of target cellular pathways for subsequent proteomic analysis. Thus, in future studies, proteomics could be used to assess protein changes in certain brain regions involved in the respiratory control network predicted by changes in genomic analysis.

In conclusion, acute ITTO conditioning in anesthetized rats elicited a load compensation response characterized by a significant increase in Te and Ttot and a decrease in Pes. ITTO also induced a change in the gene expression profile of the medial thalamus including genes involved in the stress response and anti-apoptosis. The results suggest that the medial thalamus is a component of the respiratory neural network responding to respiratory load stimuli.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).